Thick honey spreads slowly across a plate. In such a world, where friction suppresses inertia, motion is governed not by force but by viscosity. This is how many soft materials behave: polymer gels, colloidal suspensions, the contents of living cells. The researchers asked: can we locally change viscosity so that a material responds differently to stress at different points?
They developed a method based on the idea of 'directed aging': just as muscles strengthen with exercise, a material 'trains' to alter its resistance in response to local deformations. Using a mathematical approach that describes energy losses in a viscous medium, the authors derived rules for updating viscosity at each point. These rules allow the material to gradually adapt to a desired mechanical behavior.
To test the idea, the scientists simulated a disordered Maxwell material — a substance that behaves like an elastic solid under rapid forces but flows like a liquid under slow ones. They found that by locally changing viscosity, they could control the Poisson's ratio — a measure of how much a material contracts in one direction when stretched in another. For example, they could make a material that, when compressed, does not expand sideways but remains unchanged or even contracts.
This approach resembles training a neural network, but instead of adjusting connection weights, local viscosities are tuned. The material itself 'learns' the desired behavior by adapting to external conditions. This opens the way to smart gels that change stiffness in response to touch, or polymers that adapt to loads.
The Poisson's ratio of ordinary cork is close to zero: when compressed, it barely expands sideways, which is why it's easy to push a cork into a bottle.